Redesigning elastic full‐waveform inversion on the new Sunway architecture
Abstract IFOS3D is a three‐dimensional elastic full‐waveform inversion (EFWI) tool designed for high‐resolution estimation of the Earth's material properties within 3D subsurface structures. However, due to the significant computational costs associated with 3D EFWI, leveraging the computing po...
Saved in:
| Main Authors: | , , , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Wiley
2025-01-01
|
| Series: | Engineering Reports |
| Subjects: | |
| Online Access: | https://doi.org/10.1002/eng2.12819 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832576594981421056 |
|---|---|
| author | Mengyuan Hua Wubing Wan Zhaoqi Sun Zekun Yin Puyu Xiong Xiaohui Liu Haodong Tian Ping Gao Weiguo Liu Hua Wang Wenlai Zhao Zhenchun Huang |
| author_facet | Mengyuan Hua Wubing Wan Zhaoqi Sun Zekun Yin Puyu Xiong Xiaohui Liu Haodong Tian Ping Gao Weiguo Liu Hua Wang Wenlai Zhao Zhenchun Huang |
| author_sort | Mengyuan Hua |
| collection | DOAJ |
| description | Abstract IFOS3D is a three‐dimensional elastic full‐waveform inversion (EFWI) tool designed for high‐resolution estimation of the Earth's material properties within 3D subsurface structures. However, due to the significant computational costs associated with 3D EFWI, leveraging the computing power of a supercomputer for implementation is a logical choice. In this article, we introduce several innovative process‐level and thread‐level optimizations based on heterogeneous many‐core architectures in the new Sunway supercomputer, which is a powerful system globally. These optimizations encompass a process‐level communication overlapping strategy, thread‐level data partitioning and layout approaches, a remote memory access optimized master‐slave communication scheme, and a thread‐level data reuse and overlapping strategy. Through these optimizations, we achieve significant improvements in each iteration, with a kernel function speedup of approximately 59× and an overall program speedup of about 14×. Our findings demonstrate the ability of our proposed optimization strategies to overcome the computational challenges associated with 3D EFWI, providing a promising framework for future advancements in the field of subsurface imaging. |
| format | Article |
| id | doaj-art-09363bb02b1d486a9798297d51dffc41 |
| institution | Kabale University |
| issn | 2577-8196 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Wiley |
| record_format | Article |
| series | Engineering Reports |
| spelling | doaj-art-09363bb02b1d486a9798297d51dffc412025-01-31T00:22:48ZengWileyEngineering Reports2577-81962025-01-0171n/an/a10.1002/eng2.12819Redesigning elastic full‐waveform inversion on the new Sunway architectureMengyuan Hua0Wubing Wan1Zhaoqi Sun2Zekun Yin3Puyu Xiong4Xiaohui Liu5Haodong Tian6Ping Gao7Weiguo Liu8Hua Wang9Wenlai Zhao10Zhenchun Huang11School of Software Shandong University Jinan ChinaNational Supercomputing Center in Wuxi Wuxi ChinaNational Supercomputing Center in Wuxi Wuxi ChinaSchool of Software Shandong University Jinan ChinaSchool of Software Shandong University Jinan ChinaSchool of Software Shandong University Jinan ChinaSchool of Software Shandong University Jinan ChinaNational Supercomputing Center in Wuxi Wuxi ChinaSchool of Software Shandong University Jinan ChinaSchool of Software Shandong University Jinan ChinaNational Supercomputing Center in Wuxi Wuxi ChinaDepartment of Computer Science and Technology Tsinghua University Beijing ChinaAbstract IFOS3D is a three‐dimensional elastic full‐waveform inversion (EFWI) tool designed for high‐resolution estimation of the Earth's material properties within 3D subsurface structures. However, due to the significant computational costs associated with 3D EFWI, leveraging the computing power of a supercomputer for implementation is a logical choice. In this article, we introduce several innovative process‐level and thread‐level optimizations based on heterogeneous many‐core architectures in the new Sunway supercomputer, which is a powerful system globally. These optimizations encompass a process‐level communication overlapping strategy, thread‐level data partitioning and layout approaches, a remote memory access optimized master‐slave communication scheme, and a thread‐level data reuse and overlapping strategy. Through these optimizations, we achieve significant improvements in each iteration, with a kernel function speedup of approximately 59× and an overall program speedup of about 14×. Our findings demonstrate the ability of our proposed optimization strategies to overcome the computational challenges associated with 3D EFWI, providing a promising framework for future advancements in the field of subsurface imaging.https://doi.org/10.1002/eng2.12819EFWIheterogeneousHPCIFOS3DoptimizationsSunway supercomputer |
| spellingShingle | Mengyuan Hua Wubing Wan Zhaoqi Sun Zekun Yin Puyu Xiong Xiaohui Liu Haodong Tian Ping Gao Weiguo Liu Hua Wang Wenlai Zhao Zhenchun Huang Redesigning elastic full‐waveform inversion on the new Sunway architecture Engineering Reports EFWI heterogeneous HPC IFOS3D optimizations Sunway supercomputer |
| title | Redesigning elastic full‐waveform inversion on the new Sunway architecture |
| title_full | Redesigning elastic full‐waveform inversion on the new Sunway architecture |
| title_fullStr | Redesigning elastic full‐waveform inversion on the new Sunway architecture |
| title_full_unstemmed | Redesigning elastic full‐waveform inversion on the new Sunway architecture |
| title_short | Redesigning elastic full‐waveform inversion on the new Sunway architecture |
| title_sort | redesigning elastic full waveform inversion on the new sunway architecture |
| topic | EFWI heterogeneous HPC IFOS3D optimizations Sunway supercomputer |
| url | https://doi.org/10.1002/eng2.12819 |
| work_keys_str_mv | AT mengyuanhua redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT wubingwan redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT zhaoqisun redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT zekunyin redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT puyuxiong redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT xiaohuiliu redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT haodongtian redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT pinggao redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT weiguoliu redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT huawang redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT wenlaizhao redesigningelasticfullwaveforminversiononthenewsunwayarchitecture AT zhenchunhuang redesigningelasticfullwaveforminversiononthenewsunwayarchitecture |